Google Indexing API: What It Actually Works For

The Google Indexing API works for exactly two kinds of pages: those carrying JobPosting structured data, and those with a BroadcastEvent embedded in a VideoObject. Everything else — blog posts, product pages, category archives — sits outside what Google supports, whatever an "instant indexing" plugin promises on its sales page.

Google's own documentation is unambiguous. The quickstart states the API "can only be used to crawl pages with either JobPosting or BroadcastEvent embedded in a VideoObject." The reason is timing rather than favouritism: job listings and livestreams expire. A role filled on Tuesday and a stream that ended last night are worse than useless in results, so Google built a side channel for content whose value collapses on a deadline. For everything else, a sitemap and ordinary crawling already move fast enough that a separate push channel would buy very little.

Mechanically, the Google Indexing API is small. You POST to the urlNotifications:publish endpoint with a two-field JSON body — the url, and a type of either URL_UPDATED or URL_DELETED. Requests must send Content-Type: application/json. You can check what Google received with a GET to the metadata endpoint, remembering to percent-encode the address so colons become %3A and slashes %2F. Up to 100 calls can be combined into a single batch request, with each part capped at 1MB. Removal has a precondition worth respecting: the page must already return 404 or 410, or carry a noindex robots meta tag, before you ask for it to be dropped.

Quota is where the scope restriction grows teeth. New projects receive a default allowance of 200 notifications per day, which Google frames as room for onboarding and submission testing rather than a production budget. Anything beyond that requires applying for approval and resource provisioning, and that application is exactly where someone checks whether your site genuinely publishes job postings or livestream events. A site pushing ordinary articles will not get raised limits, and Google warns that attempts to game quotas across multiple accounts may result in access being revoked.

The trap that catches most people is that misuse does not fail loudly. Submit a plain blog URL and the endpoint accepts it, returns a 200, and queues it — nothing verifies at submission time that the page carries supported markup. That success response feels like proof the technique works, which is why a whole cottage industry of instant-indexing tools grew on top of it. John Mueller addressed this on Bluesky in May 2025: "We see a lot of spammers misuse the Indexing API like this, so I'd recommend just sticking to the documented & supported use-cases." Asked whether a site would be penalised, he was blunt: "I'd just use it properly, or not use it."

A second confusion sits underneath the first, and it survives even when you use the endpoint correctly. A notification asks Google to schedule a crawl; it does not buy an index entry. Pages still pass through the same quality and duplicate assessment as anything Googlebot finds by itself, so a thin page submitted through an official channel remains a thin page. Reports of URLs appearing within hours and then quietly dropping out days later are the ordinary consequence of that split, not evidence of a penalty.

So what should you use for everything else? IndexNow is the honest answer for push-style notification: an open protocol where one submission is shared with every participating engine, accepting up to 10,000 URLs per POST, with ownership proved by hosting a key file of 8 to 128 hexadecimal characters at your site root. It never reaches Google, which has not joined, but it covers Bing, Yandex and others at almost no engineering cost. For Google itself the levers are unglamorous and effective: an accurate XML sitemap, internal links from pages that already get crawled often, fast and stable responses so crawl budget is not burned on timeouts, and the URL Inspection tool in Search Console for the occasional one-off request.

The practical rule is short. If your site publishes job listings or livestream events, implement the Google Indexing API properly, request the quota, and enjoy a real advantage on freshness. If it does not, skip it entirely: install IndexNow, keep the sitemap honest, and spend the saved effort on pages worth indexing. Using an official endpoint for a purpose Google explicitly excludes is not a clever shortcut — it is a dependency on behaviour Google has said it may stop supporting without notice.